In 2016, nearly 45,000 people in the United States took their own lives, marking a 25 percent increase in suicide since 1999. Over roughly the same period, mental health expenditures in the United States more than doubled. In an especially alarming trend, the incidence of at least one major depressive episode per year among adolescents has risen by almost two-thirds over the past decade to reach 13.3 percent. Because many who experience mental health difficulties as adults first show symptoms during this developmental period, the World Health Organization lists mental health in adolescence as a top global health priority.

I have seen these trends in my own life as a professor. As a social scientist, I want to understand why. As a Christian, I wonder whether religion can help to protect adolescents from mental illness.

This has been a contentious question in the field of psychiatry. Sigmund Freud is known for his dire view of faith: “If one attempts to assign to religion its place in man’s evolution, it seems not so much to be a lasting acquisition, as a parallel to the neurosis which the civilized individual must pass through on his way from childhood to maturity.” But Carl Jung and other psychologists have taken a more positive view of the role of spirituality.

That positive effects of religiosity were strongest for the individuals presenting the most severe symptoms of depression, who are often hardest to treat.

What do statistics tell us? The preponderance of evidence supports the idea that increased religiosity is associated with better mental health. (Religiosity is measured through surveys of how often individuals attend religious services or pray, as well as how much importance they attach to religion.) But it could be that adolescents with better mental health attend church regularly and have a more stable home environment. How do we know whether it is the church attendance or the stable home environment that is benefiting them? This distinction is important: In one case the emphasis for improving mental health should be on religious involvement, and in the other, it should be on the home environment.

How could we show that religiosity has a causal effect on depression? If we could run an experiment to answer this question, we might randomly assign some adolescents to attend religious services or pray and others to avoid this activity; after a period of time, we could assess them for indicators of depression. But there are obvious ethical concerns in mandating religious attendance and otherwise convincing people to comply with the experiment. Instead, we can conduct what is called a quasi-experiment.

Students who by chance are exposed to religious peers become more religious themselves.

In collaboration with Sriya Iyer, an expert on the economics of religion at the University of Cambridge, and Anwen Zhang, an economist at the University of Glasgow, I investigated a unique data set, the National Longitudinal Survey of Adolescent to Adult Health, which includes rich questions on religiosity and depression, and we focused on the years when students were in high school.

We know that peers have strong effects in adolescence on a variety of behaviors, ranging from drug and alcohol use to academic performance. This is true for religiosity as well. When we look across grades in the same school, by chance students in some grades are exposed to more religious peers than students in other grades. This creates a sort of experiment. Students who by chance are exposed to religious peers become more religious themselves. We can isolate the effect of individual religiosity on depression using variation in individual religiosity coming solely from this random variation in the religiosity of school-grade peers. In other words, if certain adolescents become more religious as a result of exposure to religious peers, we can investigate whether these same adolescents subsequently show improved mental health.

By focusing our quasi-experiment on shifts in individual religiosity stemming solely from random exposure to peers, we are able to rule out that our estimated effects of religiosity are driven by individual background characteristics, like home environment or parental education. Because we are comparing students who are attending different grades in the same school, we can also rule out that the characteristics of the school are driving our estimated effects of religiosity.

Given that antidepressants show limited success, counselors who deal with children would be remiss to dismiss the potential beneficial effect of religiosity in treating clients.

So what do we find? Robust effects of religiosity on depression. For instance, among the students in our quasi-experiment, a 1.0 standard deviation increase in religiosity decreased the probability of being at risk of moderate to severe depression by 11 percent. (This standard deviation change in religiosity is fairly large; it is equivalent to switching from not attending church or church youth activities to attending at least once a week.) Perhaps most surprising was that these effects were strongest, almost two-thirds larger, for the individuals presenting the most severe symptoms of depression, who are often hardest to treat. This finding offers a startling contrast to evidence on the effectiveness of cognitive-based therapy, one of the most recommended forms of treatment, which is generally less effective for the most depressed individuals, at least in the short term.

Crucially, we found that the benefits of religiosity on depression are not driven merely by exposure to more religious peers in a school grade but instead by the individual student behavior. This is not to say that a more religious school environment is not helpful for the mental health of all (and this would be a great subject for future research). But all our comparisons were made across peer groups within schools to isolate the effect of individual religiosity. And we found that religiosity helps to buffer against some stressors, like worse physical health or the suicide of someone close. Furthermore, our research suggests that adolescents who have fewer support structures in place at home and in school experience more benefits from religiosity.

Interestingly, we see similar benefits of religiosity regardless of whether adolescents are active in other activities like school clubs or athletics. This suggests that these other youth activities, where adolescents can find a sense of meaning and social belonging, nevertheless do not appear to substitute for the benefits of religiosity on mental health.

Given that antidepressants show clinical success in reducing depression in only about one-fifth of cases, our research suggests that all counselors who deal with children would be remiss to dismiss the potential beneficial effect of religiosity in treating clients. With the growing body of evidence supporting a positive association between religion and mental health in many cases, research into religion is achieving growing acceptance in the field of psychiatry today, along with a variety of other fields, like economics. This is great news, as much remains to be discovered about how mental health and other important outcomes relate to faith and the inner spirit.

This article appears in September 30 2019.

Jane Cooley Fruehwirth is an associate professor in the Department of Economics at the University of North Carolina, in Chapel Hill, N.C.